Study On Fault Diagnosis System Of The Marine Diesel Based On Fuzzy Neural Network

Diesel is widely used as ship power equipment,poor working conditions, failure-prone,and complicated the process of testing,maintenance workload,once the repair and maintenance is not timely or accurate,would lead to average accident, causing property damage and marine pollution.Therefore,to do the work of diesel engine fault diagnosis is help to improve the use of diesel engine safety,reduce pollution,save maintenance costs and reduce consumption.Due to the complexity of the structure of diesel engines,it is difficult to meet the requirements for diagnosis by using the traditional method,In recent years,with the development of computers and artificial intelligence,through the use of neural networks,fuzzy logic and artificial intelligence expert system technology such as computer diagrostics is an important direction of diesel engine fault diagnosis research.This thesis has made some preliminary research on the fault diagnosis of marine diesel by using the artificial neural networks,fuzzy logic theory and expert systems theory.The main content of this thesis is as follows:(1) The research status and fault diagnosis mechanism of the marine diesel engine fault diagnosis technology was analyzed.The turbocharger system and cooling water systems have been the focus of the study.Their failure causes and the corresponding sample set were established.(2) For the complexity and ambiguity of the marine diesel engine fault,the concept of fuzzy logic has been invoked.The fuzzy neural network was constructed and was used to carry out the fault diagnosis of marine diesel engine's turbocharging system and the cooling water system.The simulation results show thatthe expression of knowledge is more accurate.It has more detailed description on the fault phenomenon and clear explanation on the cause of the malfunction,so it was fit for people's thinking habits.(3)Take advantage of the COM component technology to call Matlab Neural Network Toolbox functions for using by Visual C++ developing environment to develop the platform of the diesel engine fault expert system test platform,Fuzzy theory is used to determine the degree of fault,expert experience is used to collect the training samples and neural network is used to complete the self-learning function, which improved the efficiency of the diagnosis and overcame some shortcomings of the traditional fault diagnosis.The software has a simple and practical man-machine interface and easy to expand for the use of modular programming approach,which established the foundation of the follow-up study.